Umetna inteligenca v sodnem odločanju : magistrsko delo

Haliti, Armenda

| | SLO | ENG | Cookies and privacy

Bigger font | Smaller font

First page > Show document

Show document

Title:	Umetna inteligenca v sodnem odločanju : magistrsko delo
Authors:	ID Haliti, Armenda (Author) ID Keresteš, Tomaž (Mentor) More about this mentor...
Files:	MAG_Haliti_Armenda_2026.pdf (3,09 MB) MD5: E91BA9BB6E12EEBEE4E9E2CE8B45A03A
Language:	Slovenian
Work type:	Master's thesis/paper
Typology:	2.09 - Master's Thesis
Organization:	PF - Faculty of Law
Abstract:	V obdobju četrte industrijske revolucije, ko umetna inteligenca v veliki meri narekuje našo vsakdanjost, je bilo vprašanje časa, kdaj bo prodrla tudi na področje prava. Čeprav s tehnologijo, ki so si jo v stroki zamislili in zaželeli za človeštvo, naj ne bi rokovali še naslednjih nekaj desetletij, pa se bistvene spremembe že kažejo. Veliko prahu povzdiguje umetna inteligenca v vlogi sodnika, ideja, ki se že prakticira in stremi k prispevanju čim objektivnejših sodniških odločitev. Poleg sodniške neodvisnosti, je nepristranskost verjetno najpomembnejša konstitutivna lastnost in kvaliteta sodnika, vemo pa tudi, da na človeka vplivajo tudi psihološki in sociološki dejavniki, kar posredno ali celo neposredno potencialno obarva sprejeto sodno odločitev. Kako bi torej izgledalo, če bi namesto človeškega miselnega procesa, delo sodnika opravil računalnik? Z omembo umetne inteligence sicer še ne mislimo humanoidnih robotov s samozavedanjem, ki bi opravljali sodniško funkcijo. Govorimo o t.i. strojnem učenju, ki predstavlja analitično obdelavo podatkov na podlagi že vnešenega znanja. Gre za računalniško sposobnost avtomatiziranega prepoznavanja vzorcev, predvidevanja dogodkov in samostojnega odločanja. Osnovna problematika se pojavi že ob uokvirjanju poteka postopka sodnega odločanja, ki ga teorija široko določa in različno umešča. To torej pomeni, da vsaj v teoriji težje presodimo, kako zares poteka sodno odločanje in v katerih fragmentih sodniškega sojenja bi računalniki nadomeščali sodnike. Res pa je, da ti v praksi potrebujejo le dejanske podatke sodniških potez, ki se kažejo navzven. Z vidika pravnega varstva osebnih podatkov, predstavlja problem umetne inteligence v sodnem odločanju tudi sam postopek analize in obdelave podatkov, ki se izvede tekom učenja računalnika, ki se bo obnašal kot sodnik. Potrebna je namreč visoka stopnja profiliranja, ki povzroča polemike, prav tako so postopki samodejnega odločanja netransparentni, kar onemogoča razumevanje metodologije, ki vodi do algoritemskih rezultatov. Namen te naloge je raziskati in analizirati delo sodnikov, njihov miselni proces in dejavnike, ki nanj vplivajo. Prav tako je cilj dela spoznati osnove umetne inteligence, ratio njenega posredovanja v pravosodju, pričakovano in dosedanjo dejansko učinkovitost v vlogi sodnikov in kaj to pomeni za področje varstva osebnih podatkov, z vidika posameznika, ki je subjekt sodnikove normativne konkretizacije.
Keywords:	umetna inteligenca v sodnem odločanju, Akt o umetni inteligenci, Splošna uredba o varstvu podatkov, samodejno sprejemanje odločitev, sodniško odločanje, strojno učenje, profiliranje, analitična obdelava podatkov, varstvo osebnih podatkov, črna skrinjica, globoko učenje
Place of publishing:	Maribor
Place of performance:	Maribor
Publisher:	A. Haliti
Year of publishing:	2026
Number of pages:	1 spletni vir (1 datoteka PDF (43 str.))
PID:	20.500.12556/DKUM-96871
UDC:	340.12:004.8(043.3)
COBISS.SI-ID:	279842307
Publication date in DKUM:	29.05.2026
Views:	265
Downloads:	19
Metadata:
Categories:	PF
:	Copy citation

Average score:	(0 votes)
Your score:	Voting is allowed only for logged in users.
Share:

Hover the mouse pointer over a document title to show the abstract or click on the title to get all document metadata.

Licences

License:	CC BY-NC-ND 4.0, Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International

Link:	http://creativecommons.org/licenses/by-nc-nd/4.0/
Description:	The most restrictive Creative Commons license. This only allows people to download and share the work for no commercial gain and for no other purposes.
Licensing start date:	30.01.2026

Secondary language

Language:	English
Title:	Artificial intelligence in judicial decision - making
Abstract:	n the era of the fourth industrial revolution, when artificial intelligence largely dictates our everyday lives, it was only a matter of time before it would also penetrate the field of law. Although the technology that the profession imagined and desired for humanity is not expected to be handled for the next few decades, significant changes are already evident. Artificial intelligence in the role of a judge is raising a lot of dust, an idea that is already being practiced and strives to contribute to the most objective judicial decisions possible. In addition to judicial independence, impartiality is probably the most important constitutive characteristic and quality of a judge, and we also know that people are also influenced by psychological and sociological factors, which indirectly or even directly potentially color the adopted judicial decision. So what would it look like if, instead of a human thought process, a computer performed the work of a judge? When we mention artificial intelligence, we do not yet mean self-aware humanoid robots that would perform the function of a judge. We are talking about the so-called machine learning, which represents analytical processing of data based on previously entered knowledge. It is a computer's ability to automatically recognize patterns, predict events, and make independent decisions. The fundamental problem arises when framing the course of the judicial decision-making process, which the theory broadly defines and places differently. This means that, at least in theory, it is more difficult to judge how judicial decision-making actually takes place and in which fragments of a judicial trial computers would replace judges. However, it is true that in practice they only need actual data on the judge's actions that are visible externally. From the perspective of the legal protection of personal data, the problem of artificial intelligence in judicial decision-making is also the process of data analysis and processing itself, which is carried out during the learning of a computer that will behave as a judge. Namely, a high level of profiling is required, which causes controversy, and the automated decision-making processes are also non-transparent, which makes it impossible to understand the methodology that leads to algorithmic results. The purpose of this task is to research and analyze the work of judges, their thought process, and the factors that influence it. The aim of the work is also to learn about the basics of artificial intelligence, the rationale for its intervention in the judiciary, the expected and current actual effectiveness in the role of judges, and what this means for the field of personal data protection, from the perspective of the individual who is the subject of the judge's normative concretization.
Keywords:	artificial intelligence in judicial decision-making, Artificial Intelligence Act, General Data Protection Regulation, automatic decision-making, judicial decision-making, machine learning, profiling, analytical data processing, personal data protection, black box, deep learning

Comments

Leave comment

You must log in to leave a comment.

Comments (0)

0 - 0 / 0

There are no comments!

Back